It’s always fascinating when we begin to look at visualization across scale, both in the context of physical scale - size of representation and the impression that it leaves the viewers - as well as the contextual scale - how complex or structured the visualization is and how much good it is a procedurally displaying the hierarchy of information rather than just mindlessly laying them out. Maps are large-scale indeed presents a visually impressive visualization, which allows the creator to easily able to encode multiple layers of information in it. I definitely agree with some of the limitations and challenges raised, and want to add that there are many perceptive issues associated with the representation, that when the viewers are not able to see the full picture (literally in this case), we should question if the visualization still has the intention to present data, or has the intention switched to creating data instead. However, one thing that I do like about the real-scale visualizations is that it pushes the viewer down to a the data level, and eliminates the hierarchical bias the viewer might have when looking from a bystander perspective. The DH2019 draft was a little bit hard to understand, as the visual correlation between the multiple layers of information isn’t super clear, given the fact that the texts don’t align well with the boundaries, and the colors are more of an aesthetic choice, rather than functional. However, I am fascinated with the idea of connecting putting elevation data on top. It’s interesting that the conclusion part started with “Language is not only a means to convey one’s ideas, but also to express taste and background.”, which I feel slightly contradicts the idea that the goal of this visualization is reflect a dynamic community, given that this seems to suggest a direct correlation between visual representation, interpretation and “taste background” with pre-existing author-bias ( but it could just be me ).